Phishing Detection Using Machine Learning and Chrome Extension
Vishwanath D. Chavan, Aditya Gadekar, Sakshi Bidwai, Vivek Gogi, Laxmi Kurapati · 2024
Phishing, a common cyber threat, involves the unauthorized collection of user data through fraudulent URLs, and links. Attackers aim to persuade users to perform specific actions, enabling them to gather important information for possible unauthorized activities. This prevalent cybercrime entails spoofing trustworthy websites in order to obtain personal information like login information and passwords. Although it emerged several years ago, phishing remains a significant and ongoing threat in the cybersecurity landscape. This study presents five machine-learning algorithms designed to prevent phishing attacks by analyzing URL-based features. The proposed models, adapted to the fight against Zero-Day attacks, effectively distinguish legitimate websites from fraudulent ones. To improve user accessibility and implementation, a Chrome extension is used to conveniently and easily implement URL phishing detection. Machine learning is used in this extension to detect phishing URLs by evaluating extracted features. The research offers a thorough examination of phishing attempts and accentuates the accurateness of gradient boost Machine Learning classification methods, which is capable of reaching 96% precision, 98% recall, and 97% F1 score without the need for extra research.